This paper is the last in a series about audiological tests and test protocols (test batteries) designed to differentiate cochlear from retrocochlear siteof-lesion. While the first two papers dealt with individual tests, this paper is concerned with the audiological test protocol. A conceptual framework is presented for formulating a test protocol, and techniques are discussed for evaluating protocol performance. Using clinical data, we evaluate the performance of the traditional audiological test battery and make recommendations for the clinical use of audiological test proto-
cols.This paper is the third in a series about audiological testing for the identification of retrocochlear pathology. The first paper' presented techniques for evaluating diagnostic tests with specific application to audiology. The second paper l o evaluated the performance of audiological, vestibular, and radiological tests based upon an extensive review of the literature. This paper is concerned with the formulation and evaluation of the audiological test battery.Numerous audiological tests have been developed to help the audiologist identify the site of an abnormality in the auditory system. A fundamental principle of audiology is that these tests are never administered or interpreted in isolation, but are combined to constitute a test battery. This philosophy is espoused by many introductory or general texts on audiology. For example, in the Handbook of Clinical Audiology, p. 159,' we are told, "The audiologist neither expects nor gets complete agreement on all the different tests he performs . . . The utilization of a large number of tests, therefore, renders clarity to an otherwise muddy series of events . . . There is also a statistical advantage to the test battery approach when compared with the utilization of but a single examination finding. The multiplicity of judgments in the test battery renders the entire examination more reliable and valid." Implicit in the audiological test battery concept are three assumptions: ( 1 ) additional tests provide additional information, (2) the audiologist can effectively use this additional information, (3) the performance of the test battery 'Present address: Auditory-Vestibular Center, Panorama City, California.is always superior to the performance of individual tests. While these assumptions have a certain appeal and apparent validity, the clinical experiences of most audiologists indicate that the test battery concept is flawed. Audiologists have had the frustrating experience of using the audiological test battery without adequate information as to the best composition of the battery, the performance of the battery, or strategies for interpreting conflicting test results.The concept of the test battery has been taught and practiced for many years by audiologists without experimental validation. The evidence is insufficient to demonstrate that multiple tests offer a significant advantage over an appropriately selected individual test. Even if a test battery is advantageous, little inform...